There is no doubt that artificial intelligence, robotics and automation in the contact center offer the potential to drive faster, more consistent around-the-clock service and deliver higher customer satisfaction. Or that senior executives are enthralled with the possibilities for enhancing products and services, automating routine tasks and improving decision-making.
While organizations forge ahead with plans for AI deployments and automation projects, fewer have strategies in place to prepare the people and process parts of the equation. Uncertainty about how to get ready for the inevitable AI transformation is widespread. A series of research studies recently highlighted the issue.
At the C-level, a 2017 Deloitte survey reported that only 17% of executives are familiar with both the concept of AI and its applications at their companies. Similarly, Deloitte’s “2018 Global Human Capital Trends” report found that, while 72% of business leaders believed that AI, robots and automation are essential to increaseproductivity and performance, less than one-third (31%) felt that their organizations were prepared to address it by redesigning how work gets done and to leverage human skills like problem-solving, cognitive abilities and social skills.
Employees are just as unsure about what types of skills they’ll need in the AI-powered workplace or where they’ll obtain the additional training to avoid losing their jobs to automation. A 2018 survey by Northeastern University and Gallup found that almost half (49%) of American employees are looking to their employers to provide this training and 61% expect their employers to pay for retraining programs.
With all of this haziness swirling around AI and automation in the workplace, I decided to reach out to a few industry experts for their views on how AI is evolving the nature of work in the contact center, the impact on the human element, and advice on how to prepare your people and operation.
Read on for insights from Rick Britt, VP of Artificial Intelligence, CallMiner; Nerys Corfield, Contact Center Specialist, Unify; Fara Haron, CEO of Global BPS, Arvato; Jen Snell, Vice President, Product Marketing, Verint Intelligent Self-Service; John Thomas, Distinguished Engineer and Director, IBM Analytics, and Member, IBM Academy of Technology.
On how AI is changing the face of contact centers:
Rick Britt: We currently find two AI technologies within the contact center realm. The first is content management. This is the Watson-esque type of AI that can learn everything an organization knows and then present those pieces of information to an agent at the right time.
This marries with the second part, conversational AI, which takes a conversation between two individuals and begins to model, forecast and pattern that conversation in a way that allows you to either change the outcome of the conversation while it’s going along or derive better business strategies because of that conversation.
The initial phase of AI will be observing the patterns of these contacts and interactions. Once we have that “genome,” we can do things like optimal pathing to achieve an outcome. In addition, we can tie into that Watson system so that we can say to the agent, these are the knowledge pieces you’re going to need to learn and know during that interaction. So it will improve existing interactions.
John Thomas: AI for the call center is not just about virtual agents or appealing user interfaces. In order to be truly transformational, it needs to read the customer’s mind. It needs to understand customer intent at a deep level, be able to predict customer interactions and provide exceptional experiences. Actionable business insights come only with context. Data science/machine learning skills are needed to deliver insights that drive these experiences. We also need skills in designing interfaces that consume these insights to present compelling user experiences for a growing variety of computing environments from on-premises, to cloud environments, to the combination of both.
On which AI-powered tools deliver the most business value for improving agent performance:
Jen Snell: At a basic level, performance is improved by allowing self-service IVAs to resolve simple queries and questions, allowing for the human element without the human touch. That, in turn, allows your employees and human agents to focus on more difficult, cognitively challenging questions. For employees, the most significant way AI can help, besides answering basic questions and saving time, is the ability to improve content compliance, and time to insight as the employees no longer have to navigate 20 different screens. The employee can quickly pull up information, align answers with company policy, and field FAQs. In this way, AI-powered tools serve as an enhancement in agent performance.
On which AI applications are delivering the highest value for Early adopters:
John Thomas: Most AI early adopters are starting with training for their existing CSR staff based on insights from offline data science work on call center data. Many are also moving onto the real-time application of AI, either guiding human agents in their interactions or powering virtual interfaces that are insightful.
On key opportunities for contact centers going forward:
Nerys Corfield: In the world of contact centers, consider AI to be the 62-year-old “baby” starting to cruise and change our world. Despite AI being essentially understood since 1955, it is only now that this “baby’s” true potential is being embraced.
It is currently clear that most businesses are just moving into the cruising stage when it comes to AI. After all, AI covers a multitude of automated processing functions. From simply pointing a bot to a response library and facilitating a text-to-text “conversation” using keywords, to progressively improving decision-making through real-time sentiment analysis and heavy shifting data analytics, AI will have a definite impact on customer interactions.
Bots, through AI, are becoming commonplace and are doing a good job of taking repetitive responses away from advisors’ daily workloads. However, businesses still need to have a strong understanding of what bot limitations are and when a human assistant needs to step in and take over. It’s also important to appreciate which scenarios should never be managed by AI and will always need an injection of inherent human resources.
Key opportunities for AI, and where it is being super helpful and adopted more freely, is through the empowerment and support of advisor communities and contact centers by becoming an interactive virtual assistant that collects training tools and helps track down service matter experts. When utilized in this way, productivity levels, competency curves, eSat scores, and EFCR (effective first-contact resolution) are increasing while hold times and transfer times drop. All these embody key opportunities for contact centers going forward as they ultimately positively impact the customer experience given that they are being serviced much more effectively and efficiently.
AI, if deployed correctly, can represent an easy, frictionless experience for consumers, and an exciting opportunity for all contact center technology providers. AI is not an adjunct piece of intelligence. It should be an integral part of one’s customer journey. The use of AI will typically be an ongoing iterative process that needs to be managed by the right person within an organization. In this case, AI is not a technology solution; it is a customer experience tool and as such it should be a collaborative engagement between operations and IT with operations leading the project.
On potential challenges when developing a strategy for a hybrid contact center operation (AI and human):
Fara Haron: The term “hybrid contact center” is an ideal-state in this industry; it will be realized when we reach a sort of symbiosis between human and technological customer interaction. While this is a fully achievable ideal, as it stands today, there are several hurdles in the way—two in particular.
1. Adjusting Job Qualifications and Standards: As technologies advance so do consumers. We are seeing that the average consumer is much more tech-savvy and able to troubleshoot on their own. However, this also means that, when the consumer is not able to problem-solve alone, they will likely be reaching out for support on more complex issues that involve advanced product or process knowledge. In other words, the role of the customer service agent is becoming more advanced: it requires not only interpersonal skill, but prowess in technological comprehension and the ability to learn and re-learn quickly. For the contact center leader, this means adjusting expectations for recruitment and hiring, as well as competing for a smaller pool of applicants. On the road to becoming a hybrid operation, a major hurdle will be the competition for truly qualified candidates. Additionally, in order to attract and retain such candidates, there will likely be a need to make organizational adjustments. These might include salary raises and offering additional benefits to customer service agents.
2. Selecting and Implementing Technology Strategically: In a hybrid contact center, some activities will be fully automated with almost no human intervention and others will be a combination of human and machine. The difficulty enters when deciding what technologies to implement and when; the approach will need to be very strategic. One way of approaching these decisions tactically is by collecting data to drive decisions. Arvato makes a point to keep an eye on large-scale technologies like AI and RPA, noting their applications today, so that we can forecast the ways in which they will impact the industry tomorrow. We predict that 69% of activities in the contact center could be addressable by AI and RPA in the next decade, and this is a timeline we can keep in mind as we plan investments and implementations for the future. Still, there is a need to keep a constant pulse on these technologies, as the timeline might change.
On how to develop a digital workplace strategy that effectively integrates AI with the human element:
Jen Snell: There are three critical areas for integrating AI with humans. The first is the ability to partner across the organization. Different departments will deploy AI in different ways—some more than others. To ensure that everyone is on the same page, you need to take data out of the silos so your entire business can learn and grow from each other. Traditional contact centers are struggling to keep up with new companies that have mastered the art of AI integration.
Second, we strongly believe in data-driven use cases that take the guesswork out of where you should implement automation. AI can be used to enhance your customer experience, but it must be purposeful and create value.
Our last recommendation is to find the right vendor-partner. If you want to integrate with the human element of your business, you need to go back to those human and business needs. Finding a vendor that understands your workforce and core objectives is crucial to making sure the AI aligns with your bottom line.
On how conversational AI can be applied to improve contact center employee engagement:
Jen Snell: One of the most surprising results that conversational assistants provide for employee engagement is the value contact centers gain in strategy. With conversational assistance, we’ve seen contact centers become a source of ROI rather than just an expense. Those agents, once interchangeable, become valuable brand ambassadors and provide expertise to the organization. Not only are they in touch with customer needs, they can help craft strategies for the customer experience. So really, the value that we see is an indirect value to an employee. They suddenly become the most important pillars for strategy.
On how AI and collaboration platforms can improve team communications:
Nerys Corfield: Where AI is helpful and is being adopted more freely is through the empowerment and support of the advisor communities by becoming an interactive virtual assistant collecting training tools and finders of service matter experts. Used in this way, productivity levels and the ability to quickly address customer concerns are increasing while hold times and transfer times are decreasing. All of these things ultimately positively impact the customer experience as they are being serviced much more effectively.
AI can make several elements of daily communication and collaboration in the enterprise more productive, more streamlined, easier to manage and even smarter. At the same time, for users, it may generate a more creative and more engaging experience. Collaboration tools already make it easy to connect with colleagues via voice, video or chat, regardless of their location. Collaboration tools can also streamline communication by keeping all relevant information together, and making it easy to interact on the fly. AI can support this by empowering contact center agents or enterprise teams by acting as a virtual assistant to find information or subject-matter experts. Imagine that you are trying to figure out who are the right people (skills, knowledge, experience, availability) across your organization to answer a customer question or bring into a conference call or online meeting, and your software automatically selects and recommends them for you, based on your business context and project needs.
On top challenges facing contact centers prior to integrating AI into their operation:
John Thomas: The call center is still the most dominant channel in contact centers. Most call centers have existing systems that include IVR, call recording, agent disposition applications and basic analytics. While these systems are necessary, they present a few challenges. For starters, they tend to be fairly static, and they are often bound by closed proprietary formats. Businesses looking to integrate AI into their call center operations must be able to work with open technologies, while ensuring sensitive customer data is handled appropriately.
Rick Britt: Data will be the top challenge. It always is, it always will be. Gathering the relevant, right amount of data in a format that can be can be leveraged by machine learning is a problem in every machine learning instance. For example, if you want to predict whether a customer is going to churn, we need to know every time that you talked to the customer and what happened in every one of those conversations. We then need to find the areas that have commonalities so we can pattern those and we can predict it. However, most companies don’t have the lineage of data that they need. The majority of data within organizations is unstructured. It’s an ongoing struggle for companies.
On addressing the frontline agent “fear factor” with AI:
Fara Haron: Technology is in a constant state of change and advancement, which means that the customer service agents working with those technologies must be in a constant state of learning. In other words, communicating change is a consistent and necessary piece of managing a contact center—as such, transparency and over-communication are essential for the effort to be successful. When process changes occur in any organization, the last thing you want to do is surprise your employees suddenly. Making sure that your senior management and team leaders know your plans and are given the training and tools to regularly update their teams is essential, backed up by company updates, the intranet, or via emails.
As AI plays a more integral part in contact center work, there is also a common misconception that technology is set to replace the jobs that people currently fill. Those spearheading contact centers know this is not true, but that the role of the agent is evolving. Communication and transparency will play a key role in making this clear to agents as well. It will need to be both explained and demonstrated how the integration of new technologies into agent work is providing an opportunity to up-level the position and offering a chance to learn new skills. This understanding will help to ensure that agents’ excitement around the future of their job and entice them to stick around for the change.